pandas 计算中的最小值

Question

我想将我之前编写的一段 python 代码转换为 pandas，这样它就可以直接在 dataframe 中完成，而不是乱用 csv 文件。

我想根据多个值（属性）计算设备的健康状况。假设我有以下 df:

   A   B  C
0  7 NaN  8
1  3   3  5
2  8   1  7
3  NaN 0  3
4  8   2  7

我想计算健康度如下：

note that the def attributeHealth is still in the old form and not converted to pandas since that is the part where I get stuck and this is the code that was working with the csv library

df['Health'] = attributeHealth(df['A'], 10, 0.4) * attributeHealth(df['B'], 5, 0.5) * attributeHealth(df['C'],2 ,0.8) * 100

def attributeHealth(name, weight, limit):
            if row[name] != 'NULL':
                attrHealth = 1 - min(int(row[name])*weight/100, limit)
            else:
                attrHealth = 1
            return attrHealth

我试过先将它缩减为单个属性，但似乎我不能以这种方式使用 min():

inputDF['health'] = 1 - min(inputDF['A']* 2/100, 0.7)

提前致谢！

Answer 1

您可以为此使用 DataFrame.apply：

inputDF['health'] = inputDF.apply(lambda row: 1 - min(row['A']* 2/100, 0.7),
                                  axis=1)

apply 为每一行执行给定的可调用函数（在本例中为 lambda），returns 结果系列。

Answer 2

您可以使用 numpy.minimum 然后替换缺失值 reindex:

inputDF['health'] = ((1 - np.minimum(inputDF['A'].dropna() * 2/100, 0.7))
                            .reindex(inputDF.index, fill_value=1))

类似的解决方案：

inputDF['health'] = 1 - np.minimum(inputDF['A'].dropna() * 2/100, 0.7)
inputDF['health'] = inputDF['health'].fillna(1)                      

print (inputDF)
     A    B  C  health
0  7.0  NaN  8    0.86
1  3.0  3.0  5    0.94
2  8.0  1.0  7    0.84
3  NaN  0.0  3    1.00
4  8.0  2.0  7    0.84

总计：

def attributeHealth(col, weight, limit):
    #return Series (column)
    return ((1 - np.minimum(col.dropna() * weight/100, limit))
                               .reindex(col.index, fill_value=1))

a = attributeHealth(inputDF['A'], 10, 0.4) 
b = attributeHealth(inputDF['B'], 5, 0.5) 
c = attributeHealth(inputDF['C'], 2, 0.8) 

inputDF['Health'] = (a * b * c) * 100
print (inputDF)

     A    B  C  Health
0  7.0  NaN  8   50.40
1  3.0  3.0  5   53.55
2  8.0  1.0  7   49.02
3  NaN  0.0  3   94.00
4  8.0  2.0  7   46.44

pandas 计算中的最小值

minimum of values in pandas calculation

python

csv

min

pandas